wrote:
On 16 Apr 2006 13:52:58 -0700, "peter" wrote:

Grumman 236 wrote:
The mathematical model of the earth the the GPS software uses is
idealized somewhat and provides only an approximation of the bulging of
the surface of the earth relative to its geometric center point

But note that most consumer GPS receivers do not report altitudes based
only on the idealized mathematical model of the earth's shape
(generally the WGS-84 ellipsoid). They also include correction tables
that reflect the difference between the geoid shape and the ellipsoid
as a function of your position over the earth's surface.

Have a look at the following url:
http://sps.unavco.org/geoid/

Yes, that URL gives you the correction between geoid and ellipsoidal
heights. But as I already said, similar correction tables are already
built into most consumer GPS receivers (incl. all Garmins and
Magellans) so the altitudes they report are not what the above website
refers to as "GPS Elevation" but rather the elevation with respect to
MSL.

If you look at the NMEA data output of a GPS receiver, the altitude is
given in the $GPGGA sentence and one of the data fields specifies the
value being used by the receiver for the 'Height of geoid above WGS84
ellipsoid' when reporting the elevation.
See http://aprs.gids.nl/nmea/#gga for details.